Removing a drive without replacement.

[Esse]

Hi guys,

I see under the Dashboard that the SMART status of one of my drives shifted to red.

There is plenty of space in the array to accommodate the data from that drive without replacing it.

Is there a way to do this, make unRAID rearrange the data to the other drives in the array, and remove "the soon to be faulty drive" from the array?

 

I searched the forum and the Manual but did not find anything specific on this scenario.

 

Cheers

 

/E

[garycase]

Is the drive "red-balled" in the array?  ... or does it just have a bad SMART status?

 

If it's red-balled, it's already failed;  if it's just failing SMART, then you can indeed take quick action to get the data off and then reconfigure the array without the drive.

 

Assuming it's the latter, you can do this ...

 

(a)  Move ALL of the data on the drive to other drives on the array.  Do NOT do this using the user share references ... reference the data directly on the disk shares.    e.g.  if you have a share named "Movies" that has some data on that disk, do NOT copy it from \\Tower\Movies to another location;  copy it from \\diskX\Movies  to  \\diskY\Movies ... or to any location of your choice.

 

(b)  Once you've moved all of the data off the disk, you can then do a "New Config" and simply don't assign this disk to the array.

 

[JonathanM]

Nothing built in to unraid to do this automatically. There are some community scripts that may help, search the forum for unbalance.

 

Removing a drive will require parity to be recalculated, so you will be at extra risk of data loss until parity is built and checked without that drive.

 

Make sure the rest of your drives are healthy before you start moving a bunch of stuff around.

 

As always, making your backups current should be a high priority.

[garycase]

There are actually ways to do this without requiring recalculation of parity ... i.e. writing all zeroes to the drive while it's still parity protected will then allow a "New Config" with the "trust parity" option [i.e. you can check the "Parity is already valid" box].

 

But it's far easier for most to just do a New Config and a new parity sync.

 

... just remember, as jonathan noted, that you're going to be running "at risk" (no fault tolerance) until the new parity sync is completed, so if there are files you don't have backed up that you don't want to lose, it's a good idea to be sure your backups are up-to-date before doing this.

 

 

[Esse]

Thanks guys!

Yes, just a bad SMART status.

Will this work:

1/ Move the content of the "faulty" drive over to the remaining drives.

2/ Take the array offline, shutdown and remove the drive.

3/ Boot up, run the "New config" (scary, have never done this before  ), will this only see the drives that are left in the array and give them new numbers?

4/ Take the array online and run a parity check.

 

In "New config" it says: "This is a utility to reset the array disk configuration so that all disks appear as "New" disks, as if it were a fresh new server."

Does this mean that all data on the rest of the drives gets wiped?

 

Cheers

 

/E

[garycase]

No, it doesn't "wipe out" any data on the drives.

 

DO be sure you know (a) which drive is the parity drive; and (b) which drive(s), if any, are assigned as cache drives.

 

The easiest way to be certain you assign the drives you want is to print a copy of the WebGUI page that shows all of your current drive assignments => then when you do the New Config you can be sure you assign them correctly.    It doesn't really matter which "slot" you use for which data drive ... but it DOES matter than you assign the correct drive as parity, as that one WILL get completely rewritten when you do the new parity sync.

 

[Esse]

Aha, ok, then it should not be a problem.

 

This was great info guys, just in case I go for this scenario.

 

Thanks again!

 

/E